Databricks Release Notes

MLflow trace storage in Unity Catalog tables will be available by default for workspaces with the compliance security profile enabled in mid-August 2026.

Store MLflow traces in Unity Catalog tables for scalable, governed trace storage: storage is unlimited, traces are stored in OpenTelemetry (OTel) format, access is governed through Unity Catalog schema and table permissions, and you can query traces with SQL. To view traces, point an MLflow experiment at a Unity Catalog trace location.

See Store OpenTelemetry traces in Unity Catalog.

• Search and reuse past conversations: You can now search your past Genie One conversations and bring previous context into your current conversation. See Recall past conversations.
• Memory (Beta): Genie One can remember details you ask it to save, such as your preferences and workflows, and applies them in future conversations. See Add to Genie One's memory.
• Mention scheduled tasks: You can now @mention scheduled tasks in Genie One. See Scheduled tasks.
• New chat loading fix: Fixed a bug where a new chat showed a blank screen while waiting for the first response.

Register Scala and Java user-defined functions (UDFs) in Unity Catalog and reuse them across notebooks, jobs, and SQL warehouses with Unity Catalog governance. Reuse existing JVM logic and get better performance than Python UDFs on all compute, including serverless.

See Scala and Java user-defined functions (UDFs) in Unity Catalog.

For workspaces with the compliance security profile enabled, see What's coming for planned availability changes.

MLflow trace storage in Unity Catalog tables is now generally available. Databricks recommends storing traces in Unity Catalog for new and production workloads: storage is unlimited, traces are stored in OpenTelemetry (OTel) format, access is governed through Unity Catalog schema and table permissions, and you can query traces with SQL, Genie, AI/BI dashboards, and any Spark-based tool. To view traces, point an MLflow experiment at a Unity Catalog trace location. For supported regions, see Features with limited regional availability.

See Store OpenTelemetry traces in Unity Catalog and Observe and find issues.

The agentic code converter is in Beta. Use it to convert legacy SQL dialects, such as T-SQL, Snowflake, Redshift, Oracle, BigQuery, and Teradata, to ANSI SQL in migration projects in your workspace. Powered by Genie Code, the converter analyzes each source script, validates syntax and semantic intent, and iteratively fixes errors before writing the results to your workspace. A workspace admin must enable it from the Previews page.

See Convert SQL with the agentic code converter.

Google Gemini 3 Pro is retired. Gemini 3.1 Pro is the recommended replacement. See Deprecated and retired models for migration guidance.

AI Search high QPS will be available by default for workspaces with the compliance security profile enabled in late August 2026.

Set a target QPS on standard AI Search endpoints to power high-throughput, real-time workloads like search bars, recommendation systems, and entity matching. High QPS is enabled by default, but existing endpoints are not affected. Additional capacity and the associated cost apply only when you configure a target QPS for an endpoint.

See Scale AI Search endpoint throughput with high QPS.

Genie One and Genie Agents usage is free through January 31, 2027. Budget controls do not apply during this time.

Lakebase now provides a Service Level Agreement (SLA), with Monthly Uptime Percentage commitments for endpoints in zone-redundant high availability mode and for single-compute endpoints. For details, see Lakebase SLA.

Model Serving now supports Thinking Machine Labs Inkling as a Databricks-hosted model. This mixture of experts (MoE) model supports text and image inputs, a 1 million token context window, and is optimized for coding, reasoning, and agentic tool-use tasks. You can access this model using Foundation Model APIs.

Inkling is available in Public Preview.

Customers are responsible for ensuring their compliance with applicable terms.

Data Classification is now available by default for workspaces with the compliance security profile enabled and either C5, TISAX, or IRAP controls selected.

You can now claim and enable a Custom URL that gives your Databricks account a single, branded entry point (for example, acme.databricks.com) for all users and workspaces. See Give your Databricks account a custom URL.

AI Search High QPS is now generally available for standard endpoints.

Set a target QPS on standard AI Search endpoints to power high-throughput, real-time workloads like search bars, recommendation systems, and entity matching. High QPS is enabled by default, but existing endpoints are not affected. Additional capacity and the associated cost apply only when you configure a target QPS for an endpoint.

See Scale AI Search endpoint throughput with high QPS.

Query-based connectors in Lakeflow Connect will soon be available by default for workspaces with the compliance security profile enabled. Query-based connectors ingest data from databases by querying the source directly using a cursor column, without requiring change data capture (CDC) configuration or an ingestion gateway. Supported sources include Oracle, Teradata, SQL Server, MySQL, MariaDB, and PostgreSQL, as well as all Lakehouse Federation data sources using foreign catalog ingestion.

See Query-based connectors.

In an upcoming release, Request for Access will be enabled by default for all workspaces, and access requests will be routed to the owner of the requested Unity Catalog object via email. You can configure access request destinations at any time in the metastore settings. The legacy workspace-level access request settings will be deprecated.

See Enable access requests across the metastore.

Lakebase is now SOC 2 Type 2 compliant. For our SOC reports, please ask your Databricks account team or see Compliance.

Context-based ingress control now supports Partner Platforms as a network source. Select a partner platform to allowlist the IPs that third-party apps use to connect to Databricks, including Power BI, Tableau Cloud, and dbt platform. Databricks manages and updates these IP lists automatically. See Network sources.